汽车行业游戏规则改变者的新动力

The New Driving Force of theAutomotive Industry Authors & Contact person Lead Augustin FriedelSoftware Defined VehiclesAugustin.Friedel@mhp.com Author Marcus WillandMobilityMarcus.Willand@mhp.com Lead Matthias BorchArtificial IntelligenceMatthias.Borch@mhp.com Author Dr. Nils SchaupensteinerTransformation AdvisoryNils.Schaupensteiner@mhp.com Contact Person Stephan BaierArtificial IntelligenceStephan.Baier@mhp.com AuthorPatrick RuhlandTransformation AdvisoryPatrick.Ruhland@mhp.com The study “AI as Game Changer“ and its summary were published by:MHP Gesellschaft für Management- und IT-Beratung mbH All rights reserved! No reproduction, microfilming, storage, or processing in electronic media permitted without the consent of thepublisher. The contents of this publication are intended to inform our customers and business partners. Theycorrespond to the state of knowledge of the authors at the time of publication. To resolve any issues, please referto the sources listed in the publication or contact the designated contact persons. Opinion articles reflect theviews of the individual authors. Rounding differences may occur in the graphics. Contents ContentsTable of figures12 Key FindingsWelcome to Change! 01. Revolution and Automotive Market Potential11 03. Pilot Projects and Implementation19 04. AI Models, Levels, and Use Cases 23 4.1 The Game Changer: What Can Be Achieved With AI264.2 Automobile Manufacturers With Low AI Investment294.3 AI Models: Make or Buy?29 05. AI Applications Along the Automotive Value Chain31 5.1 AI Operation in Vehicles and in the Cloud355.2 AI Monetization in Vehicles395.3 Added Value of AI Applications in Companies40 06. What the Customer Wants: The User Perspective 47 6.1 Use and Understanding of AI Applications496.2 Advantages and Disadvantages – Generally and in Vehicles496.3 Purchasing Decision, Trust and Willingnessto Pay51 07. Success Factors and Strategic Approach 55 7.1 Strategy and Goal Planning567.2 Think from the Perspective of the Customer, not the Technology567.3 Organizational Anchoring and Ownership587.4 Local Differences require local Setup597.5 Reducing Complexity597.6 Use andMonetization of Data607.7 Checklist for successful Implementation61 08. Challenges, Responsibility, and Risks63 8.1 Costs of Trainingand Operation648.2 Data andDigitalization as a Basis658.3 Business Models and Cases for B2C and B2B658.4 Ethics and Responsibility678.5 New Risks and Regulatory Challenges69 09. AI Applications in the Automotive Industry: 7 Recommendations for Action71 10. Further Informations 75 Literature and SourcesContactInternationalAboutMHP 78 79 Table of figures Figure 1:Technology super cycles – artificial intelligence as the next relevant platform shift(Coatue, 2024)12Figure 2: AI market size in the automotive sector (Precedence Research, 2024)12Figure 3:Total investments in AI companies founded since 2001, in USD billion (Scheuer, 2024)16Figure 4: Investment in AI stack layers (Coatue, 2024)17Figure 5: Companies with team and budget for AI (Capgemini, 2023)21Figure 6: Interconnected AI concepts24Figure 7: Visualization of AI as a pyramid25Figure 8: Classification of AI terms27Figure 9: The performance of AI models compared to human capabilities in the MMLU test (iAsk, 2024)28 Figure 10: Schematic diagram of the training of AI foundation models for vehicles30Figure 11: Use of AI along the value chain32Figure 12: Significant improvements of functions and features through AI33Figure 13: Interest in AI functions compared internationally34Figure 14: Role of on-premise, cloud, and vehicle for AI models35Figure 15: Levels of a software-defined vehicle (SDV) (Willand, Friedel, & Schaupensteiner, 2023)36Figure 16: Different models for ADAS and AD applications and functions37Figure 17: AI’s potential at different stages of the value chain(Capgemini, 2023)40Figure 18: Use of AI-based solutions by region41Figure 19: Key drivers behind the use of AI in production42 Figure 20: Decisive issue – fewer users of software due to AI or free software (Coatue, 2024)43Figure 21: Possible uses of AI in software development(Wee 2024)44Figure 22: Understanding of AI in cars48Figure 23: Advantages of using AI in cars49Figure24: The perceived advantages and disadvantages of using AI50Figure25: AI in cars: purchase motivation or blocker?51Figure 26: Trust in stakeholders with regard to the implementation of AI in vehicles52Figure 27: Willingness to pay for AI functions52Abb. 28: Assessment of the future AI competence of car manufacturers by region53Figure29: Customer and use case first, and then AI applications and models57 Figure 30: Dimensions for validating technical feasibility57Figure 31: Training costs for AI models are increasing (Stanford University, 2024)64Figure 32: Data availability and quality by region65Figure 33: Customers’ willingness to pay is unclear; costs arise for implementation and operation66Figure34: Classification of AI use case categori